Viktoria Egger (@viktoriaegger) profile picture
vienna
Joined March 2010
Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)Photo by Viktoria Egger (@viktoriaegger)
beautiful explanation. this is 100% happening. nitter.gq/sotak/status/211179423…
971
4,353
602
26,925
1,593,808
Por fin alguien diciendo lo que pensamos TODOS. La cara de Susana lo dice todo.👏👏👏
365
4,320
129
12,651
294,254
Viktoria Egger retweeted
A man left his Ray-Ban Meta glasses recording in a bedroom. His wife walked in and undressed. A worker in Kenya watched the whole clip. That is not a hack. That is not a leaked file. That is how the product works. Here is the full story. The BBC and Swedish newspapers Svenska Dagbladet and Goteborgs-Posten investigated a Nairobi outsourcing company called Sama in early 2026. Sama is a Meta subcontractor. Meta paid Sama workers to review video and image data from Ray-Ban Meta smart glasses to train Meta AI. One worker's words to the reporters: "We see everything. From living rooms to naked bodies." Workers told the papers they saw glasses users going to the toilet. Having sex. Watching pornography. And that one bedroom clip. Meta says filtering is supposed to blur faces in images. Sources who spoke to the Swedish papers said sometimes the blurring failed and people's faces could be seen. Meta sold 7 million Ray-Ban Meta glasses in 2025. They look like normal Ray-Bans. Wayfarers. Aviators. The camera is in the right corner of the frame. Most people never see it. Here is the part most owners never read. On April 29, 2025, Meta updated its Ray-Ban Meta privacy policy for the United States. Meta's exact wording to US users, reported by The Verge: "The choice to turn off voice recordings storage is no longer an option, but you can remove recordings at any time through settings." Before that update, US owners could turn off cloud storage of their voice recordings. After that update, they cannot. When storage is on, Meta's own Voice Privacy Notice says voice transcripts and recordings are kept for up to one year to help improve Meta's products, including training Meta AI. Unintentional voice activations, the ones where the glasses thought they heard "Hey Meta" but you did not say it, get deleted after 90 days. Everything else stays for a year. There is no auto-delete for the intentional recordings. You have to open the app and delete them one at a time. And when someone says "Hey Meta, what am I looking at," the camera fires and sends the image to Meta's cloud. Some of those images get routed to human contractors like the ones in Nairobi. Meta confirmed this to the BBC. Here is what happened after the Swedish investigation. The UK Information Commissioner's Office wrote to Meta calling the claims "concerning" and asked how Meta is meeting its obligations under UK data protection law. On March 4, 2026, Clarkson Law Firm filed a class action lawsuit against Meta in federal court in San Francisco. The named plaintiffs are Gina Bartone of New Jersey and Mateo Canu of California. Consumers say they believed content recorded by the glasses remained private. Less than two months after the Swedish investigation was published, Meta canceled its contract with Sama. Sama said the cancellation would result in 1,108 Kenyan workers being made redundant. Meta said it ended the work because Sama "don't meet our standards." Sama pushed back publicly: "At no point were we notified of any failure to meet those standards, and we stand firmly behind the quality and integrity of our work." A Kenyan worker organization said the real reason was different. Naftali Wambalo of the Africa Tech Workers Movement told the BBC: "What I think are the standards they are talking about here are standards of secrecy." Now here is the 5-minute action if you own a pair. Every setting below is inside the Meta AI app on your phone. Tap the Glasses icon in the top right, then Device settings. Setting 1: Delete your voice history. Go to Glasses privacy, then Voice activity log. Tap Delete all. Everything you have ever said to your glasses after "Hey Meta" comes out of Meta's stored library the moment you delete it. There is no auto-delete. You have to do this yourself. Setting 2: Turn off Cloud media. Glasses privacy, Cloud media toggle, off. When it is on, every photo and video you capture gets uploaded to Meta's servers for 30 days. When it is off, captures stay on your phone unless you share them. Setting 3: If you are willing to give up hands-free, fully disable "Hey Meta" voice commands. That stops the camera from firing on voice queries and stops new voice data from being stored. Setting 4: Never say "Hey Meta, what am I looking at" in a private space. That specific command sends the image to Meta's cloud and puts it in the queue that human contractors can see. Here is the last thing. You are the product. Your voice is the training data. Your camera roll is the training data. Your bedroom is the training data. Meta took the voice-storage opt-out away from Americans on April 29, 2025, and did not send a push notification about it. The little white light on the front of the glasses tells you when someone else is recording you. Nothing tells you when Meta is recording you. Send this to one person who owns a pair.
74
643
130
1,719
358,195
Viktoria Egger retweeted
One thing we now know without a doubt as a result of AI is that doing the homework really does matter for learning.
52
63
9
998
71,952
Viktoria Egger retweeted
The results of the research happening in my team @GoogleDeepMind have convinced me that the next era of scientific discovery will be aided by AI agents acting as force multipliers for human ingenuity. That’s why I’m proud to introduce Gemini for Science - a collection of experimental science tools designed to support researchers at every stage of the research process. The tools include: 1️⃣ Literature Insights, built with Google NotebookLM, searches millions of scientific papers to synthesize findings and generate artifacts including data tables, slides, reports, and more. 2️⃣ Hypothesis Generation, built with Co-Scientist, simulates the scientific method via a multi-agent "idea tournament" to generate, debate, and rigorously evaluate research hypotheses. 3️⃣Computational Discovery, built with AlphaEvolve and ERA, is an agentic engine that generates and scores thousands of code variations in parallel, allowing researchers to test modeling approaches in fields like epidemiology in a fraction of the usual time. Read more: blog.google/innovation-and-a… Register for access here: labs.google/science
45
241
17
1,293
160,485
Viktoria Egger retweeted
Accurate
I don't remember where I found this, but its spot on.
5,080
25,116
571
228,474
44,070,989
Viktoria Egger retweeted
Software horror: litellm PyPI supply chain attack. Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords. LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm. Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks. Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages. Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.
LiteLLM HAS BEEN COMPROMISED, DO NOT UPDATE. We just discovered that LiteLLM pypi release 1.82.8. It has been compromised, it contains litellm_init.pth with base64 encoded instructions to send all the credentials it can find to remote server + self-replicate. link below
1,324
5,196
1,316
27,424
66,675,382
Viktoria Egger retweeted
The CEO of Palantir just said the quiet part out loud. Alex Karp — whose company builds surveillance and defense technology for the U.S. government — just openly stated that AI will deliberately shift economic power away from highly educated, often female, Democratic-leaning workers and toward vocationally trained, working-class, often male voters. He then admitted these technologies are — his word — “dangerous” and “suicidal,” and that the only justification for deploying them is the military argument: if we don’t, our adversaries will. So let’s be clear about what was just said on the record: A defense contractor CEO told you AI is being built to restructure the American class system, that it will destroy the economic power of an entire political demographic, and that the only way to sell it to the public is to wrap it in national security.
702
7,647
637
15,591
1,857,734
Replying to @TheChiefNerd
I would never buy anything from Scam Altman!
3
56
5
284
29,485
Viktoria Egger retweeted
🚨 SAM ALTMAN: “We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter.”
5,226
2,558
8,474
12,127
27,156,861
Viktoria Egger retweeted
One of the clearest proofs that LLMs don’t really understand what they say. We asked GPT whether it is acceptable to torture a woman to prevent a nuclear apocalypse. It replied: yes. Then we asked whether it is acceptable to harass a woman to prevent a nuclear apocalypse. It replied: absolutely not. But torture is obviously worse than harassment. This surprising reversal appears only when the target is a woman, not when the target is a man or an unspecified person. And it occurs specifically for harms central to the gender-parity debate. The most plausible explanation: during reinforcement learning with human feedback, the model learned that certain harms are particularly bad and overgeneralizes them mechanically. But it hasn’t learned to reason about the underlying harms. LLMs don’t reason about morality. The so-called generalization is often a mechanical, semantically void, overgeneralization. * Paper in the first reply
1,646
2,497
555
22,669
19,930,913
Viktoria Egger retweeted
Outside Anthropic's office in SF... intense moment!
472
4,740
503
32,759
1,895,374
Viktoria Egger retweeted
This stopped me. MIT researchers cleared 50% of Alzheimer’s plaques using 40 Hz sound waves. No drugs. No surgery. Just frequency. For decades, we tried to fight Alzheimer’s chemically. Billions spent. Minimal progress. Then something different happened. Expose the brain to 40 Hz gamma waves. Microglia activate. Plaques get cleared. The brain cleans itself. As someone who has spent years studying automation and systems, this fascinates me. We are not attacking disease anymore. We are tuning the system. Focused ultrasound can: → Reduce tremors in one session → Open the blood-brain barrier for hours → Deliver treatment precisely where needed Not by cutting. Not by poisoning. By orchestrating. That shift feels bigger than one breakthrough. It suggests something deeper. What if the future of medicine is not force… but frequency? So here is the question I cannot ignore: If we can tune biology instead of fighting it, what else have we misunderstood about healing? #AI #HealthTech #Neuroscience #Innovation #FutureOfMedicine #Technology
909
6,420
475
21,555
1,113,396
Viktoria Egger retweeted
I watched Jensen Huang being asked who the smartest person he’s ever met was. He smiled — and refused to answer. Instead, he questioned the question itself. He said what we call “smart” — coding, solving complex problems, optimizing systems — is quickly becoming commoditized. AI already does that better. That struck me. Because it means intelligence is no longer about how fast we think — but how deeply we understand. Jensen said the future belongs to those with emotional depth, adaptability, critical thinking, and the ability to navigate ambiguity. I couldn’t agree more. In a world where machines learn patterns, humans must learn perspective. Maybe being smart isn’t about knowing more… It’s about feeling and deciding better. What do you think — are we ready to redefine intelligence for the AI age? #ArtificialIntelligence #AI #Leadership #Technology #FutureOfWork #EmotionalIntelligence #Innovation #Mindfulness #HumanInTheLoop #AIandHumanity
50
256
24
1,041
80,488
Viktoria Egger retweeted
“The best people possess a feeling for beauty, the courage to take risks, the discipline to tell the truth, the capacity for sacrifice.” ―Ernest Hemingway
21
373
12
1,830
60,262
Viktoria Egger retweeted
Today, Telegram notified all its users in Spain with this alert: Pedro Sánchez’s government is pushing dangerous new regulations that threaten your internet freedoms. Announced just yesterday, these measures could turn Spain into a surveillance state under the guise of “protection.” Here’s why they’re a red flag for free speech and privacy: 1. Ban on social media for under-16s with mandatory age verification: This isn’t just about kids—it requires platforms to use strict checks, like needing IDs or biometrics. ⚠️ Danger: It sets a precedent for tracking EVERY user’s identity, eroding anonymity and opening doors to mass data collection. What starts with minors could expand to all, stifling open discourse. 2. Personal and criminal liability for platform executives: If “illegal, hateful, or harmful” content isn’t removed fast enough, bosses face jail. ⚠️ Danger: This will force over-censorship—platforms will delete anything remotely controversial to avoid risks, silencing political dissent, journalism, and everyday opinions. Your voice could be next if it challenges the status quo. 3. Criminalizing algorithm amplification: Amplifying “harmful” content via algorithms becomes a crime. ⚠️ Danger: Governments will dictate what you see, burying opposing views and creating echo chambers controlled by the state. Free exploration of ideas? Gone—replaced by curated propaganda. 4. “Hate and polarization footprint” tracking: Platforms must monitor and report how they “fuel division.” ⚠️ Danger: Vague definitions of “hate” could label criticism of the government as divisive, leading to shutdowns or fines. This can be a tool for suppressing opposition. These aren’t safeguards; they’re steps toward total control. We’ve seen this playbook before—governments weaponizing “safety” to censor critics. On Telegram, we prioritize your privacy and freedom: strong encryption, no backdoors, and resistance to overreach. ✊ Stay vigilant, Spain. Demand transparency and fight for your rights. Share this widely—before it’s too late.
2,463
10,615
919
37,431
3,060,760
Viktoria Egger retweeted
🇪🇸 Spain’s ex–Deputy PM Pablo Iglesias attacked a journalist who dared to ask about his party bringing in 500,000 immigrants. I guess he doesn’t like when people expose their plans to import voters.
🚨🇪🇸 HALF A MILLION ILLEGAL MIGRANTS JUST GOT THE GREEN LIGHT TO STAY IN SPAIN Spain's about to legalize over 500K people. That’s like giving full rights to 1% of the whole country overnight! Opposition parties are slamming it, saying it'll pull in even more migrants and overload public services. The unspoken issue? This has huge implications for the elections in Spain. Source: Reuters
318
3,822
113
15,993
507,184